Two shift operation combined cycle power HRSG operation challenge

1 Combined cycle power plant operation mode

Depends the power plant located the place, some power gird has different engergy source wind power , coal power, neclear power, solar panel, which will make the combine cycle power plant  operation mode changed, opeation can be two shift operation or base load operation

In the meaning time the gas turbine cooling method  also impact the operation flexilability of gas turbine, steam cooling gas turbine requests longer time warm up steam cooling system , also request a auxiliary boiler  operaion to provide the steam for gas turbine cooling,  it is take longer time  to start up  steam cooling gass turbine  then air cooling gas turbine.

 For the  combined cycle power plant also has signal shaft and also 2×1 and 3 x1 power plant configuation, single shaft use less land than combined  multipe shaft, single  shaft also can start faster then  multiple shaft.

Combined cycle power plant develop very fast at the early stage 21 century, which based on plenty on the nature gas source, high efficency and less evnviromental reless, the technology provide gas turbine delveop fast and efficiency can be reach over 60%, and gas turbine capacity also increase. The build time is short then another competitor. Combustor outlet temperature can be reach over 1700 degree C, the better combustion control technoloty make the NOx generation be controlled

Ontario, Canada installed necular power plant, the nuclear power plant can’t to be the picking power plant, and wind solar generation will not stop if there is sorce enrgeny to generate electrity, at night time the system total capacity will be half of the total demand, so the all gas turbine power plant required by the dispatch center to stop, the two shift operation can’t be avoid.

SGEC signed contrat  with AESO regards the two shift operation mode, AESO paid  the start up gas and and mainternace fee, if the power plant running at 70% based load at day time, then AESO would not pay to power plant, anytime if the power plant dispatch to 100% and also request to duct burner operation, the protion which is above 70% would get  paid by the AESO, so the plant relabiliaty is important to make sure to catch  the peaking load.

2  two shift combined cycle power plant face chanllege

Two shift combined cycle power plant face a lot chanllege compare with based load power plant, the start up frequency can cause the thermal cycle, HRSG from cold to  hot start, the heat surface expaned more than 200 mm expension, and heat surface tube welding face heating cycle and very easy to broken up,

at the time author worked power plant which as two shift combined cycle experienced two shift operation, 300 time start up per years, never have chance to operation the power plant pass the over night, so the HRSG heating surface  broken many times due to the frequency start up, and the the drain pipe line peneration bellow damaged  brorken frequecy, the ciruclation pump motor broken and cooling tower gear box broken many times  which was not installed soft start device and cooling tower distrubution pipe broken due to the water hammer

Two  shift start up also  can cause more demin water useage, the demain water will have more chanellenge than based load plant

Siemens  501F gas tubine uses start up motor to start up gas turbine which damaged once due to frequency start

Frequency start up can cause the start up silencer working harder than normal, the theremal shock will can cause the silencer crack and if the it used the cooling tower water as the cooling source which can cause the STG silsencer damage due to calcuim scale to plug the diffusion pipe small holes.

During  start up time if the steam system drain not properly, it can cause steam hammer damaged the hanger system.

Frequency start can  cause the gas turbine exhanust cylinder and exhaust strut crack and increase the vibration of gas turbine

For the cooling tower if the ciruclaiton and cooling tower fan not install the soft start divice,  frequency start can cause the motor fauilure and repaire cost will increase,cooling tower no de-icing function to cause icing on winter time(figure 0)

Figure 0  Cooling tower icing build on winter time

Another point is  plant design aspect, the plant designer maybe not enough experience to design the plant runing two shift operation mode, so the equipment selection was not consider enough, for example the SGCE plant, the orignial designed for running the plant at south of the USA, so the design for the site weather condition is not fit up operation   

The gas turbine not design the inlet heating system and there was no snow hood for the air cooling generatior inlet and outlet strainer, when the first winter came, heavy snow covered the inlet to cuase the plant shutdown, the gas turbine generator lead bus peneration panel no weather roof, the heavy rain and sonw which very easy to move through to cause millinon dollar lost of three phase short circuit.

The plant had long team service agreement with Siemens, which paid to siemens for eacy start up then Semens proviode outage service and parts to support the power plant opeation.  

3 HRSG heat surface face the challenge at two shift

3.1 Drum operation challenge

HRSG installed 3 drums, HP drum, IP drum and LP drum, two shift operation plant HP drum faces change was the thermal cycle, HP drum has the thickest wall if the heating fast, the temperautre difference gredient will be bigger which can cause the stress on the HP drum downcomer nozzle and HP drum welding specially the lognitiudal welding, so the OEM will provide a temperature ramp up speed control  cruverto prevent stress welding and cause the crack.

3.2  Heating surface

Heating surface bundle face challenge, the tube will is far thinner than header, the two shift operation plant, when the gas turbine start up, the exhaust gas temperature increase very fast, the HP superheater and reheater tube will rapidely grow up to push the header gowing down, some tube may  get more heating than another one will grow faster, that will cause the extra stress than other tube, if the stress level is over the maximum tensile strees, the tube possible crack can be broken.

SGEC experienced multiple time the HRSG tube broken due to frequency start up and stop, the symptom of the tube broken was  the make up water quantity increase, sharp noise from inside of the HRSG bundle, and if the tube broken, then lost the water will cause drum level can’t maintanin ether the feed water pump flow reach maixum.

Some time when unit have time stop, visually using flash light to check the bottom of the tube, author at least find two time the tube leaking, using flash light careful review the each tube if there is sign  show very cleaning surface, the oxidant material on the tube surface become polished, then there is small  tube leaking , watch each tube color at the header welding heat effective zone, which is the maximum stress on it , most time the tube broken was happened at that zone.

Repair the tube also has challenge due to the accessable, some tube was most far location, which can’t access the tube then requested to cut the surronding tube to access the damage tube, the tube with fin long enough should be pushed it  back then allow room to access the tube, after repaired, the inspection  required use Dye peneratrion or use MT, some time use RT maybe faster.

Repiar plan need to report locatl authority inpsector to get approve, and the weld prorcedure  need be approve, before repaire, plant engiener need to let contractor to understand the tube  matreial, welder qualification need to be checked.  The tube repair was stright forward, actually the ASME codes not require RT or UT test due the diameter was samll.

After repair the local authority inspector required serive test before inspector to sign the repaire documents.

3.3  Serveral leakage and repair picture introduce

3.3.1 HRSG#1 hot reheater tube leakage

Figure 1 is showing the HRSG#1 Hot reheater  tube leakage, the leakage was on the heat effective zon form the thick welding to transfer to tube thickenss, this area had highest stress accumulation when the fast start up or cooling, the broken tube expension faster than another tube thermal stree cause the tensile stress over heated , repair was not diffecut which leakage just in the front of the tube bundle at the duct burner bay.

Figure 1 is showing the HRSG#1 Hot reheater  tube leakage

3.3.2 HRSG#2 IP evaporator leakage

Figure 2  is showing the HRSG#2 IP evaporator leakage at the header welding, the welding area always had more chance than another place, welding defact, theremal stress. Those kind of leakage most likely not regards FAC corrosion due to the temperature is over 300 degree C. the leakage repair is not difficult based on it is not need to cut extra tubes.

Figuer 2 HRSG#2 IP evaporator leakage

3.3.3 HRSG#3 HRH  superheater top header tube broken

HRSG #3 hot reheater top header find one tube broken after the unit shutdown, this header is behind of the HP superheater 1, so need to cut the HP superheater 1 to access the broken tube, to access the upper header to utilze the skycliamber to support the tube repair, to repaire one tube leakage, require cut 18 tube to support repair.

 Figure 3A HRSG#3 HRH  superheater top header tube broken

Figure 3B HRSG#3 HRH  superheater top header tube broken repair

3.3.4 HRSG#2 HRH reheater lower header two tube broken

Figure 4,5 is showing HRSG #2 heat reheater lower header found two  lower header tube broken at same time, so requestedb to cut the HP steam superheater three row to access the fauilure tubes   and completed the repair, the repair is relevant easy the top header, the header need skyclimber each time only 2 people can access, which cause the longer time to repair and test.

Figure 4 HRSG#3 HRH  superheater top header tube broken repair

Figure 5 HRSG#3 HRH  superheater top header tube broken repair

3.3.5 HRSG 1 HRH tube broken at south side panel lower header

HRSG#1 HRH tube find one broken, this tube located on the south side header not require to cut the tube but need to cut a access window at the HRSG insulated wall  for  the repair .

Figure 6 HRSG 1 HRH tube broken

3.3.6 HRSG#2 HRH north panel tube broken

HRSG#2 find one of HRH tube broken which is similar position of the HRSG#1, to access the broken tube required to cut the HP steam superheater first header three row tubes to access.

Figure 7 HRSG#2 HRH north panel tube broken

3.3.7 HRSG #3 HP steam superheater leakage

Figure 8 leaking was HP supper header the first row lower header has a crack to leakage the condensate after the unit shutdown and first chance to do inspection, so immediately inspection the lower tube is important after unit down, and not open the drain before inspection, condensate accumulation in the lower header was help to inspection, if open the drain early, the leakage will not easy to find, repair is easy, not need to cut any tube. after inspected open the drain immediately to prevent the pit corrosion at inside of the tubes. The repair pross is most easy one author did at that time, no cut tubes, no skyclimber, not need scaffolding just beside of the access door!

Figure 8 HRSG #3 HP steam superheater tube leaking

3.3.8 HRSG #1 HRH  steam superheater tube leaking

Another tube broken was found at HRSG 1 HRH  superheater but at different location , the leakage is not easy access need to cut side window of HRSG, the broken on the weld effect zone.

Figure 8A HRSG #1 HRH steam superheater tube leaking

Figure 8B  HRSG #1 HRH  steam superheater tube leaking

3.3. 9 HRSG#1 HRH tube broken root cause analysis

One broken piece was sent to metallurgist for the root cause analysis, below is digest for the report

Figure 9 the broken tube for the root cause analysis

  • Root cause general

The tube, located at the end of superheater header #1 (south bank), fractured by a dissimilar metal weld (DMW) failure. All of the macro characteristics of DMW failures were observed. Fracture occurred in the T23 tube along the weld fusion boundary, by a highly localized creep mechanism. Carbon depletion and softening of the lower alloyed material in this narrow zone is a characteristic of DMW failures.

The stresses caused by operation, especially during thermal transient periods, were the driving force for the fracture. Additionally, it is believed there were extraordinary stresses due to different conditions at the extreme end of the superheater headers. These are thought to have been caused by desuperheater overspray, and by unintended water flow in the outer tubes during upset conditions. Both of these conditions exacerbate thermally-induced stresses at the tube-to-header joints.

The tube composition conformed to all revisions of the applicable standard and Code Case. Hardness was in the lower end of the normal range. Minor surface decarburization and coarse carbonitride precipitates point to tube processing deficiencies; these are undesirable but are not implicated in the failure.

  • Summary of the founding for the broken piece

A single 1.5″Ø x 0.205″ tube fractured at the weld connecting it to the upper superheater header. The weld attached to the header appeared intact, and no visible ductility or secondary cracks were apparent to the unaided eye. The essentially brittle fracture occurred along the fusion boundary of the weld, but within tube material. This was confirmed by metallographic examination. Oxide covered the entire fracture surface, except where it had spalled off. Based on comparison with the oxide on the tube surfaces, it is estimated (very approximately) that the tube has been fully broken for up to 21% of its service life, and that it was partially broken and leaking significantly earlier. Oxide thickness varied over the fracture surface. The east side had the thickest scale, suggesting that it separated earliest. The thickness was lowest on the opposite, west side.

A section of tube remote from the fracture and outside of the bend zone was examined to establish a baseline for structure and properties. Chemical analysis confirmed that the composition met the requirements of Code Case 2199, in the original and all subsequent revisions. The core structure was consistent with tempered bainite, with ample decoration of grain boundaries by carbide precipitates. V,Nb(C,N) carbonitride precipitates were noticeable, and indicate possible tube processing deficiencies.

The tube exhibited mill decarburization to a depth of 0.25 mm (10 mils) at the inner surface and 0.20 mm (8 mils) at the outer surface. Core hardness of the T23 tube was 170-200 HV, toward the lower end of the normal range. The degree of tempering, as indicated by both hardness and structure suggest

that post-weld heat treatment temperature was fairly high. The low hardness values in the tube HAZ also support this conclusion.

  • Root cause

The differential metal combination at the weld is by definition a necessary precondition for DMW failure. The root causes are materials, structural design, and loads induced by operating conditions, and their interactions. the modified 9% chromium filler metal (B9) selected for these welds is overmatched relative to the tube, both in terms of creep strength and alloy content1. At high temperatures, chromium attracts carbon and reacts with it to form carbides. Carbon diffusion occurs during post-weld heat treatment, but also more slowly at the service temperature. Despite similar carbon contents in the weld and tube, chromium causes ‘uphill diffusion’ of carbon across the fusion line. This depletes the tube of carbon in a narrow zone along the fusion line and enriches the higher-chromium weld metal. Carbon depletion significantly lowers the creep strength in a narrow zone. Compounding the problem is the inherently low creep ductility of low alloy Cr-Mo steels, particularly under conditions of triaxial stress. The softening effect was not conclusively found in this failure, but all of the other identifying characteristics of DMW were observed.

Transferring the alloy transition (and the potential for DMW failure) to the header side of the joint moves it to a location where the metal thickness is much higher, and where stresses from steam pressure, bending, and differential thermal expansion (all drivers of DMW fracture) are lower. Superheater tube-to-header joints are at a critical location for all types of loading, and  are subject to significant stresses from each of the above sources.

 The structural design of the equipment, and particularly the very high header:tube thickness ratio of more than 8:1, makes the tube-to-header joints very prone to stresses caused by differential expansion during thermal transient periods. The tube heats up much more rapidly than the header at start-up, and vice versa during ramping down at the end of the cycle; the consequent differential diametral expansion causes high cyclic stresses, and is a characteristic of DMW joints. Stresses due to bending moments are a maximum at the tube ends, just before the transition to thicker material. Additional stresses result from

temperature variations along the header and between tube locations.

Different conditions found at the extreme end of the header are thought to have contributed to preferential failure there. Steam flow may be lower at the outer tubes (less cooling), while combustion gas velocity between the outside tubes and the sidewall is greater (more tube heating), so that thermal transients are more severe and steady state temperature is higher.

First fracture on the East side implies that bending loads favoured the vertical east-west plane (normal to the header axis), and that tensile stresses were greatest on that side of the tube. The ramping down at the end of the cycle and desuperheater overspray can cause the tube to be cooler than the header, resulting in bending moments acting on the header connections for the two outer rows (that use bent tubes). This scenario is consistent with first fracture on the East side.

There was communication with the HRSG manufacturer, who mentioned that desuperheater overspray has been implicated in superheater tube failures. Another comment was “… it has been found that water tends to be forced to the outer tubes during system upset conditions and can contribute to low cycle fatigue/tensile overload failures in ‘odd’ places”.


4  HRSG Siesmeic  bumper expode

Siesmic bumper designed by HRSG manufacture to install a bumper on the each side HRSG module to prevent the HRSG horizonal moving due to siemic , for example the earth quick happen. It was used a 18 inch pipe and sealed on both side, it was a small vessel, but designer not thank about this, the pressure at the HRSG gas is low, but is is experied several bumper explode. As figure 10 was showing the air inside of the bumper heated and incrased pressure to explode the bumper, the broken pieces to damaged the side baffle, the cut the tube fins and damaged one tube, was required forced outage to repaire tube, baffle and bumper. So after this explode, plant cut a hole to all the bumper release the pressure as figure

Figure 10 HRSG seismic explode

Figure 11 broken bumper piece damaged side baffle

Figure 11 broken bumper  cut the fins and damaged the tube  

Figure 12  cut a hole to all the seismic bumper   

5 Feed water pump cast iron leakage

Another leaking out of the HRSG tube is the feed water pump cat iron body pin hole leakage Figure 13A,B

Figure 13A pin hole leakage at feed water pump body

Figure 13B pin hole leakage at feed water pump body

6 HRSG expansion bellow broken

When frequency start up the old design of the expansion bellow broken frequency, and repair to new style expansion bellow, as figure below, the new style expansion joint design form HRST installed and have good results, plant scheduled to replace them all to new style to over come this problems.

Figure 14  expansion bellow broken

Figure 15, drain line attached with the supporter and bending due to design issue

Figure 16A expansion bellow damaged

Figure 16B   expansion bellow damaged

Figure 17 A   new style expansion bellow installed

Figure 17 B   new style expansion bellow installed

Figure 18    expansion damaged pipe supporter

Figure 19 expansion damaged the hanger

7  HRSG hot reheater drain line broken frequency

HRSG HP steam and Hot reheater drain design is not reasonable which not enough expansion bending pipe, it was broken several times, redesign it using more bending pipe line on the bottom to absorb the reheater and supper heater module expansion, it  worked well.

Figure 20 economizer drain pipe leakage

Figure 21 Hot reheater drain pipe

Figure 22 HP steam superheater drain pipe broken

Figure 23 HP steam superheater drain pipe broken

Figure 24 HRH steam superheater drain pipe broken

Figure 25 HRH steam superheater drain pipe new designed to avoid broken due to expansion

Figure 26 kettle boiler expansion joint leakage due to very thin wall

8 Conclusion

Two shift operation HRSG had more problem then based load operation plant due to frequency start up and shutdown cause the thermal cycle, this paper lists several pictures to demonstrated two shift operation power plant has more challenge and designer for the power plant need to consider the situation the plant future operation to help power plant success down the road.

Leave a Comment

Your email address will not be published. Required fields are marked *